Policy Analysis in South Africa
with
Regional Applied General Equilibrium Models
M J Cameron
Dissertation submitted in (partial) fulfilment of the requirements for the degree Master of Commercii in economics
at the Potchefstroom Campus of the North-West University
Supervisor: Prof W A Naude
Abstract
The research question addressed in this study is first to illustrate the appropriateness and feasibility of a Regional Applied General Equilibrium (RAGE) model in the South African context for policy makers. Secondly the study provides proposals for the construction of a RAGE model to assist policy-makers to investigate spatial or regional (sub-national) impacts and policy-choices with a more sophisticated approach (other than a simplified top-down approach or partial equilibrium models) which would provide for more consistent results or implications as related to the spatial or regional implications of such impacts.
This study is an attempt to address this shortcoming by investigating and illustrating the need for such modelling ability. This is done by illustrating the value and potential of RAGE models by making use of a simplified "top-down" regional application. In particular, the shortcomings of such a 'top-down' approach are highlighted and the case made for a more sophisticated 'bottoms up' approach.
A literature review was conducted regarding the types of RAGE models that can be found, as well as on the various methodological approaches to regional or spatial Applied General Equilibrium (AGE) modelling. A simplified top-down model based on the AGE model used by South Africa's Industrial Development Corporation (IDC) was constructed in order to conduct some illustrative applications. These illustrative applications highlighted some of the shortcomings of a top-down approach to regional AGE models based on an application of an electricity price and volume shock scenario.
Finally, this study presents proposals for future research regarding the development of RAGE models for the South African economy. These take account of the current work on Provincial Social Accounting Matrices (SAMs) lead by the Development Bank of Southern Africa. The feasibility of a hybrid regional AGE was illustrated and some future potential developments in this regard were stated.
JEL Classification: D58, 132, N70, N77, O20, 025, 029, 055, Q48
Keywords: Applied General Equilibrium Model, Computable General Equilibrium Model, Energy policy, Electricity, Economic Development, South Africa.
Summary
One of the legacies of the policies of the apartheid regime manifested in the South African economy today is the uneven economic development of South Africa's sub-national economies. In order to achieve greater geographical equity a specific focus on the geographic spread of economic activities and investment has been adopted by the government. There are various challenges emanating from the policies and strategies and ultimately the government needs to be able to improve the coherence of these strategies and ensure a more coordinated implementation, as well as some way of identifying sub-national (regional) trends and being able to evaluate the potential impact or implications of potential policy decisions, while also being in a position to monitor the effects of such policy implementations on the targeted regions.
Access to modelling tools with a regional or sub-national dimension enabling policy makers to analyse potential impacts of proposed strategies or test "what-if scenarios would complement the policy makers' arsenal of resources to apply to these questions. Through a chronology of Applied General Equilibrium (AGE) modelling developments based on Rossouw (2004), this study has focused on illustrating that AGE modelling tools addressing this spatial or regional dimension are limited in the South African context. Currently, it is the author's opinion that there exists no such model of sufficient regional detail and sophistication in order to conduct really useful analysis of this nature.
This study contributes to an attempt to address this shortcoming by investigating and illustrating the need for such modelling ability, while also illustrating the value and potential of such analysis by making use of a simplified "top-down" regional application based on a version of the Industrial Development Corporation's (by now dated) AGE model - IDCGEM.
A plethora of application opportunities for such a model currently exists and more will be uncovered once such a tool exists and is applied. Some examples were mentioned and a key topical example was used for the illustrative application of the IDCGEM - that of the current electricity shortage in South Africa. The subsequent issues of proposed price increases and lower supply availability related to this electricity shortage will have diverse impacts on the different provinces in South Africa. The illustrative example served to highlight some of the shortcomings of a top-down approach to regional AGE analysis.
One of the key issues or constraints to constructing such a regional AGE highlighted in the study is the existence and availability of suitably detailed inter-regional data sets to inform such a model. Through this research the author has also illustrated the feasibility to construct a more pragmatic (due to data constraints) hybrid (mix between pure top-down and bottom-up) model with the potential to be realised in the short to medium term based on the work (lead by the Development Bank of Southern Africa) currently being conducted to construct provincial social accounting matrices (SAMs) for South Africa for the base year 2006.
The recommendation from the study is that further work would need to be conducted in order to realise such a model. This would entail consolidating and extending the provincial and national SAMs into a coherent database suitable for use with a regional AGE model. In this case the Monash ORANI-style modelling would be proposed, as a lot of previous experience and knowledge with this specific approach have been gained in South Africa over the last 15 years. This approach has also been applied in various other countries for regional AGE models.
Once this model is operational and has been tested, relevant policy issues can be analysed to provide insight and assistance to policy makers with regards to sub-national/provincial implications of various policies.
Finally this study aimed to illustrate the need, viability and requirements in the South African context for a new, current regional AGE model. The actual development of such a regional AGE is left for future research.
Acknowledgements
This dissertation has been in the making for nearly ten years. I just never made the time to sit down and do it! I would like to thank my wonderful wife for allowing me to hog time from her and the children to complete this long overdue dissertation. I would also like to thank my parents for enabling me to pursue academic studies in the first place - without their assistance I would not be where I am today.
The other two people who made this happen that I sincerely would like to thank is Prof. Wim Naude, who got me interested in the topic in the first place. This was a first for South Africa when we started way back in the beginning of 1993. In the meantime life happened and it got delayed. However, the star that just doesn't give up on one is Prof. Wilma Viviers, who just would not stop pushing and pulling until I finally had to give in! Thank you Wilma, for all the motivation and trust. One last word of thanks is to Dr. Mark Horridge from Monash University, Australia. Mark taught me how to speak GEMPACK -QED!
I wish to acknowledge with thanks the financial support that this dissertation has received from the National Research Foundation (NRF).
Martin Cameron May 2008
Centurion
CONTENTS
Chapter 1: Introduction 1 1.1. Purpose of this study. 1
1.2. Hypothesis 1 1.3. Background and problem statement 1
1.3.1. Background 1 1.3.2. Problem statement 4 1.4. Research question 7 1.5. Objectives 8 1.6. Methodology 8 1.7. Terminology 8 1.8. Layout of the study 10
Chapter 2: A theoretical overview of AGE models 12
2.1. Introduction 12 2.2. The need for AGE models 13
2.3. General equilibrium 15 2.4. General application of AGE models 16
2.4.1. Methodological advantages 18 2.4.2. Policy applications advantages 19 2.4.3. Methodological disadvantages 20 2.4.4. Policy applications disadvantages 21 2.4.5. Model application disadvantages 21 2.5. Basic building blocks for AGE models 24
2.5.1. Input-output analysis 24 2.5.2. Social accounting matrices 25 2.6. The AGE framework. 26
2.7. Closure rules for AGE models 29 2.8. Interpretation of AGE results 30
2.9. Summary 32 Chapter 3: Applied general equilibrium modelling in South Africa 34
3.1. Introduction 34 3.2. Major applications of AGE models in South Africa 35
3.3. Focus of AGE models in South Africa 38
3.4. Summary 40 Chapter 4: Regional applied general equilibrium modelling 41
4.1. Introduction 41 4.2. Regional development challenges in South Africa 41
4.3. Approaches to Regional AGE modelling 43
4.3.1. The top-down approach 45 4.3.2. The bottom-up approach 45 4.3.3. The hybrid approach 46 4.3.4. Summary of approaches 46 4.4. Examples of regional A GE model applications 46
4.5. General relevance of RA GE model applications 52 4.6. South African specific relevance of RA GE model applications ....54
4.7. Summary 56 Chapter 5: Illustration of a top-down RAGE model for South Africa 57
5.1. Introduction 57 5.2. Description of the IDCGEM 58
5.3. IDCGEM regional disaggregation 62 5.4. Illustrative scenario - simulating electricity load shedding 67
5.4.2. Model closure applied and other assumptions 69 5.4.3. Analysis of the national level impacts of the scenario 72
5.4.4. Analysis of the regional level impacts of the scenario 74
5.5. Suggested future extensions 82 5.6. A pragmatic hybrid approach 87
5.7. Summary. 92 Chapter 6: Summary, conclusions and recommendations 94
5.8. Summary 94 5.9. Conclusions 95 5.10. Recommendations 97
Annexure A - GEMPACK Source code for regional extension oflDCGEM 99
LIST OF TABLES
Table 1: Chronological developments of AGE modelling in SA 1993-2005 ...37
Table 2: International example applications of RAGEs 49 Table 3: Comparison between regional contributions estimates - IDCGEM &
StatsSA 64 Table 4: IDCGEM provincial sectoral shares 65
Table 5: Regional local versus national industries 67
Table 6: National level impacts 72 Table 7: Regional 46 sector results 77 Table 8: Activities included in the North-West SAM 91
LIST OF FIGURES
Figure 1: Simplified relationship between and IO table, SAM and AGE Model 28 Figure 2: Stylised representation of change reported in AGE model results..31
Figure 3: Stylized range of regional AGE model approaches 44 Figure 4: Example geographic representation of output from the spatial AGE
model for Tokyo 50 Figure 5: Example geographic representation of output from the CGEurope
model 51 Figure 6: IDCGEM provincial shares of GDP 63
Figure 7: Example of energy specific model technical adjustments 71
Figure 8: Regional GDP impacts of scenario 75 Figure 9: Regional Employment impacts of scenario 76
Figure 10: Electricity usage by customer category 80 Figure 11: Extensions to production function required for a bottom-up RAGE
model 84 Figure 12: Extensions to production function required for a hybrid bottom-up
CHAPTER 1: INTRODUCTION
1.1. Purpose of this study
This study aims to illustrate the relevance, potential and need for an applied Regional Applied1 General Equilibrium (RAGE) model for South Africa.
1.2. Hypothesis
The hypothesis underlying this study is that a RAGE model can be of use to better understand the economic challenges faced by South African policy makers on a sub-national (regional or local) level. These challenges include (as will be discussed in more detail below) growing spatial inequalities, spatial peculiarities and idiosyncrasies in resources endowments, and different sub-national impacts of sub-national policies and intersub-national trends.
1.3. Background and problem statement
1.3.1. Background
Why is there a need for applied general equilibrium (AGE) models? There are at least three reasons that can be given to answer this question:
(a) to improve the understanding of economic processes; (b) to make predictions; and
1 In literature on the topic of general equilibrium models one finds references to both "Applied"
and "Computable" general equilibrium models used interchangeably. The prior refers to the
broader concept of the theoretical framework applied in practice, while the latter more
specifically refers to the actual model applied on a computer with a specific piece of software
such as GAMS (General Algebraic Modeling System) or GEMPACK (General Equilibrium
(c) to analyse the likely effects of economic policy and changes in such policy.
For reasons (a) and (b) many models have been constructed, and are more or less effective under different conditions. For reason (c), the options are more limited and the choice in any given case will depend for example on the type of policy being investigated, the number of sectors affected, the relevant time period, and a few other factors.
In the context of the questions posed above, AGE models have provided unique insights into the mechanics of economies and on the possible impacts of various macroeconomic policies. AGE models represent a major enhancement in economic analysis. However, one still has to be cautious with the application of decisions based on a AGE model analysis, as it is not yet clear how accurate AGE models are quantitatively, particularly in comparison to other types of models and despite many success stories in the field of AGE modelling, there is a need for something more in terms of surety regarding the AGE relationship with reality (Iqbal and Siddiqui, 2001:22).
Despite their limitations AGE models are becoming more widely used. The advent of the micro-computer has allowed the application of AGE models by policymakers and academics in policy analysis to proliferate since the late 1970s (Bandara, 1991:10). These developments have provided unique insights into the working of economies and on the potential effects of various macroeconomic policies. Due to the fact that constructing, implementing, testing, and maintaining an AGE model for any economy is a major task and
resource intensive; only limited portions of individual careers have been devoted to this field of study and modelling. Institutional effort and funding is normally required, both for developing theory as well as enabling applications of such theory in applied models. Joint research programmes of modellers, econometricians, theoreticians, and computational economists might be necessary to sustain an effective large scale AGE model of an economy (McDonald & Punt, 2005:90).
Internationally various policy issues have been addressed by making use of AGE models over the last 35 years. In South Africa though, AGE models only started receiving attention from applied quantitative economists since the early 1990s. AGE models make use of a general equilibrium framework through which the feedback and trickle down effects of an economic policy question being studied is reflected, while using a consistent database and typically needing data for only a single year (Naude and Coetzee, 2004:915). An additional advantage of AGE models is that they are explicitly grounded on micro- and macroeconomic theory and can therefore avoid accusations of potential "data mining" or "a theoretical time-series analysis" (Sims, 1980:15; Hendry, 1980:14; Darnell and Evans, 1990:113).
"Regional" and "spatial" applications of AGE models were a natural next level of development for AGE models. While the data requirements are more onerous than for national-level-only models, the benefit of being able to analyse implications of policy or other impacts on a more disaggregated level
within the economy outweighs the cost of compiling this additional information.
Leontief et al. (1965) developed a basic methodology to achieve a first level of "regional" disaggregation by dividing industries into national and local industries. National industries are defined as those industries that typically produce commodities or products that are readily traded between regions (e.g. agricultural, mineral and non-perishable manufactured goods and some services such as government services). Local industries in turn are defined as industries producing perishable goods and services that are not readily traded between regions.
By applying the above approach and through further refinements and further research, various approaches exist today to incorporate both multi-regional and spatial dimensions into AGE models. See e.g. Parmenter et al. (1985), Parmenter et al. (2000) and Adams et al. (2003).
1.3.2. Problem statement
One of the legacies of the policies of the apartheid regime manifested in the South African economy today is the uneven economic development of South Africa's sub-national or regional economies (Bond, 2005; Curtis, 1984; Morris and Robbins, 2004; Naude and Krugell, 2003; 2006).
In order to achieve greater geographical equity a specific focus on the geographic spread of economic activities and investment has been adopted by the government, and more specifically the Department of Trade and Industry (thedti). A range of policies are already in place in this regard, such
as the Integrated Sustainable Rural Development Strategy (ISRDS), the Urban Renewal Strategy (URS), Industrial Development Zones (IDZs), Spatial Development Initiatives (SDIs), and the approach of Integrated Development Plans (IDPs) (Cameron, 2005:146). Furthermore, local government is viewed as a key catalyst for locally-led social and economic development by the
Department of Provincial and Local Government (DPLG) and so-called Local Economic Development (LED) strategies and plans are required for all municipalities in South Africa (Bond, 2002).
There are various challenges emanating from these policies and strategies and ultimately the government needs to be able to improve the coherence of these strategies and ensure a more coordinated implementation. At the same time government also need some way of identifying sub-national (regional) trends and be able to evaluate the potential regional impact or implications of potential policy decisions, while also being in a position to monitor the effects of such policy implementations on the targeted regions. In this regard, policy recommendations and the implementation of resulting programs or projects have to be linked to areas of high poverty and unemployment and a more geo-technical approach to both data and analysis is required in order to satisfy these analysis and monitoring needs (Cameron, 2005:146).
The old statement that "in order to go somewhere you need to know where you are first' comes to mind in the context of this issue. Both national and local government has a need to be able to assess the demographic, economic and socio-economic status quo, as well as measure growth and development
on a sub-national level and more specifically on magisterial district, local metro and municipal levels. Indeed, governments on provincial and local level have a need for information on the spatial implications of their policies due to their constitutional obligations to promote economic development in their jurisdictions (Jansen van Rensburg and Naude, 2007; Naude, 2005).
Information and estimates are needed for activities such as: policy and strategy decisions;
economic planning; market development; and
for infrastructure planning, development and delivery
Therefore there exists a real need for more, timely as well as better quality sub-national (regional or provincial) and sub-provincial (spatial, municipal or magisterial) information to be collected, collated and made available in an easy to access manner to decision makers, than what is currently available from formal sources.
In addition to data and information, access to modelling tools with a regional or sub-national dimension that enables policy makers to analyse potential impacts of proposed strategies or test "what-if scenarios would complement the policy makers' arsenal of resources to apply to these questions. However, modelling tools addressing this spatial or regional dimension are limited in the South African context. Currently, to the author's knowledge, there exists no such model of sufficient regional detail and sophistication in order to conduct useful analysis of this nature.
This study contributes to an attempt to address this shortcoming by investigating and illustrating the need for such modelling ability. The study also illustrates the value and potential of such analysis by making use of a simplified "top-down" regional application, while at the same time highlighting some of the shortcomings of a simplified top-down approach to regional AGE modelling in the South African context. The illustrative application is based on a version of the Industrial Development Corporation's (IDC) (by now dated) applied general equilibrium (AGE) model - IDCGEM.
This study therefore aims to illustrate the need, viability and requirements in the South African context for such a new, current regional AGE model. The actual development of such a regional AGE is left for future research.
1.4. Research question
The research question to be addressed in this study is firstly to illustrate the appropriateness and feasibility of a regional AGE model in the South African context for South African policy makers.
In addition to the initial aim, the study also forwards some thoughts and ideas on how a regional AGE model can be constructed to assist South African policy makers to investigate spatial or regional (sub-national) impacts and policy choices with a more sophisticated approach (other than a simplified top-down approach) which would provide for more accurate and informative results or implications as related to the spatial or regional implications of such impacts.
1.5. Objectives
The objectives set for this study therefore are the following:
a. To illustrate the appropriateness of RAGE model as a modelling framework for regional / spatial or sub-national policy analysis.
b. To identify various ways and methodologies in which a RAGE model can be constructed in the South African context.
c. To apply a simplified top-down RAGE model to illustrate its usefulness based on South African topical issues.
1.6. Methodology
An international literature review was conducted regarding the types of RAGE models that exist and are applied in practice internationally. The literature review also served to inform on the various methodological approaches to regional or spatial AGE modelling (e.g. top-down, versus bottom-up approaches). A simplified top-down model based on the existing IDCGEM model was applied in order to conduct some illustrative applications of a RAGE model for the South African context. This illustrative application highlighted some of the shortcomings of a top-down approach to regional AGE models based on an application of an electricity price and volume shock scenario.
1.7. Terminology
An Applied General Equilibrium (AGE) model can briefly be described as an economy-wide model that includes feedback between demand, income and production structure, and where all prices adjust until decisions made in production are consistent with decisions made in demand (Dervis et al.,
1985). These models are also referred to as Computable General Equilibrium (CGE) models, especially in the case where a practical computer-based model has been constructed.
"Regional" in this context refers to a sub-national regional dimension to the
model. For the purpose of this research we prefer to use the term regional, however in the literature and types of models there is no consistency of application of "regional" in the context of intra-country versus inter-country applications of the term. In the context of this research we refer to inter-country applications as "multi-inter-country".
The term "spatial" in the context of this study is applied to refer to "micro" or "localised" applications, such as e.g. just a specific large city or metropolitan area.
Furthermore, in the context of this study in general "regional" refers to a provincial level dimension. However, this could also be expanded into lower levels of detail to for example magisterial, municipal or even large city level detail as for example in the case of Melbourne, Australia (see Horridge,
1999).
In the context of regional analysis in the specific case of South African political and legal geography it is important to understand the relationship or context between provincial, magisterial and municipal demarcations and concepts for South Africa. De Lange (2005) states that the South African government
consists of three distinct, but interrelated spheres of government, namely the national, provincial2 and local spheres of government.
The building blocks of South Africa's local sphere of government consist of municipalities and differ significantly from magisterial districts. Magisterial districts are defined areas in which South African courts exercise jurisdiction, whereas municipal boundaries are the geographical areas, designated by the
Municipal Demarcation Board, within which a democratically elected legislative and executive authority exercise the powers granted to such municipality in the Constitution of South Africa and subsequent laws.
1.8. Layout of the study
The rest of this study is presented as follows:
• In chapter 2 a brief overview of the nature of AGE models is given.
• Chapter 3 provides a historical overview of the development and use of AGE models in South Africa. From this chapter it is concluded that there is currently a lack of RAGE modelling in South Africa.
• Regional AGE models and their relevance for South African policy makers are illustrated in Chapter 4. In this chapter the focus is on understanding how RAGE models are constructed and used elsewhere
2 The provincial sphere of government was established based on the pre-1994 nine provinces
of the Republic, out of the areas of the former Republic of South Africa, and the Bantustans of
Transkei, Bophuthatswana, Venda and Ciskei and the areas of the six self-governing
territories. The geographical areas comprising each of the nine provinces is and still are
defined by reference to specific magisterial districts created in terms of the Magistrates'
in the world, and to identify the generic components of a typical RAGE model.
• In chapter 5 a RAGE model is illustrated for South Africa. Here, a simplified "top-down" application is illustrated using the IDC's AGE model, which is known as IDCGEM.
• Finally, chapter 6 contains a summary of the study and conclusions, with recommendations for future research,
CHAPTER 2: A THEORETICAL OVERVIEW OF AGE MODELS
2.1. Introduction
The previous chapter provided the background, problem statement, and motivation for the study as well as the methodology to be applied and delimitation of this specific study. It was stated that access to modelling tools with a regional or sub-national dimension enabling policy makers to analyse potential impacts of proposed strategies or test "what-if scenarios would complement the policy makers' arsenal of resources to apply to these questions. The main research question posed in the previous chapter to be addressed in this study is to illustrate the appropriateness and feasibility of a regional AGE model in the South African context for South African policy makers.
In order to put the discussion of the regional application of AGE models in context, this chapter provides a theoretical overview of the nature of AGE models. The chapter is structured as follows. In section 2.2 the motivation for the development of AGE models are set out. In section 2.3 an explanation of the concept of general equilibrium, is given. In section 2.4 the general application of AGE models is described, while in section 2.5 AGE models and their relationships with Input-Output (IO) analysis as well as Social Accounting Matrices (SAMs) are discussed. The AGE conceptual framework and concept of closures are discussed in sections 2.6 and 2.7 respectively, followed by an explanation of the interpretation of AGE percentage change results in section 2.8. Lastly, section 2.9 summarises.
2.2. The need for AGE models
The challenge the economic policy maker at all levels of government faces (as well as private business institutions operating for profit) is that limited resources must be allocated to areas that are most likely to achieve success in scenarios with the greatest probability of being realized. In order to assist and inform policy makers in this task economic models play an important role (Cameron, 2005:148).
"Economic theory does make unrealistic assumptions. .. But some abstraction from reality is necessary because of the incredible complexity of the economic world. .. Abstraction from unimportant details is necessary to understand the functioning of anything as complex as the economy."
(Baumol and Blinder, 1997:9).
Economic models are mathematical representations of the economy that are designed to be simplifications of a complex reality. They combine those behavioural relationships believed to be responsible for the bulk of macroeconomic fluctuations, while omitting those deemed less important. This process of differentiation allows economists to make predictions that are reasonably accurate and that can be more easily understood and communicated to policymakers and other stakeholders (Coletti and Murchison, 2002:20).
Models can provide key perceptions for analysis of comprehensive packages of economic and non-economic policy instruments within a consistent
framework. Insights generated by modelling can help in evaluating various policy options. At the same time, however, policy options which seem technically and even economically feasible at the sectoral level may lose their attraction when the policy maker discovers their potential effects on the economy as a whole, or vice versa (Cameron, 2005:148).
Business risk, sector-specific institutional barriers and market imperfections can all frustrate the economic process. When they exist, they certainly will not disappear by themselves. Policies thus should aim to create a framework within which more economic opportunities become market possibilities. Such policies require knowledge (which models can help to provide) of the extent to which these obstacles endure and can be removed cost effectively (Coletti and Murchison, 2002:20).
In the longer term, technological evolution, as well as unforeseen changes in consumer behaviour and preferences complicates any economic impact assessment. Ignoring the long-term impact of short-term decisions cannot be cost effective except by accident. The dynamic challenge of sequential decision-making is to link short-term decisions and long-term goals with enough flexibility to cope with the uncertainties (Coletti and Murchison, 2002:21).
Finally, economic models help settle debates that cannot be settled by theory alone. Economic theory often suggests that potentially offsetting influences are at work in the economy. When combined with statistical methods, models
help economists quantify the relative importance of each factor, thereby providing an estimate of the net impact of these offsetting influences (Cameron, 2005:149).
2.3. General equilibrium
General equilibrium refers to a state in an economy where the needs of all participants in this economy are satisfied. This implies that there exists no excess demand for, or supply of, any goods or services traded in this economy (Arrow, 1974: 254).
The formalisation of the concept of general equilibrium is commonly attributed to Walras (1874) and his publication "Elements of pure economics". Walras observed that prices were formulated simultaneously in all markets in an economy. By using a mathematical system of simultaneous equations to describe the interaction between all participants involved in the economy, Walras attempted to prove the existence of a vector of prices that would allow the system to solve, but was unable to do so (Arrow and Debreu, 1954: 265). As a result of his endeavours, Walras developed what today is commonly referred to in Economics circles as Walras' law. This law states that if there are M markets in an economy, and (M-1) of these are in equilibrium, it implies that the M-th market will also be in equilibrium (Weintraub, 1977:4).
The actual formal proof of the existence of such a price vector and state of equilibrium was only supplied in 1954 by Arrow and Debreu (Shoven and Whalley, 1992:12). Further developments of the theory of general equilibrium analysis focus on the existence and uniqueness of equilibrium. For more
detail see Arrow and Debreu (1954), Debreu (1959) and Hahn (1973). General equilibrium modelling evolved further based on the groundwork laid by Leontief (1936), linear programming models (LPs) and GAP-models and the path-breaking work of Johansen with his Multi-Sectoral Growth model in the 1960s (See Balderston, 1954; Dorfman, et al., 1958; Chenery and Strout, 1966; and Bacha, 1990, for more information on these developments).
The dynamic nature of the economy and markets dictates that it is impossible to determine a condition where equilibrium will exist. However, an estimation for the equilibrium condition is attained by "freezing" the economy at a specific time and performing the calculations based on the values at that specific instance in time by using an applied general equilibrium model (AGE).
The AGE modelling approach is an empirical counterpart of the theoretical "general equilibrium analysis" mentioned above (Bandara, 1991:5). The aim of AGE modelling is to convert the abstract representation of an economy into realistic, solvable models of actual economies. In the AGE framework the main focus of analysis is quantitative and is based on the empirical data from a particular country being investigated. One of the major features of AGE modelling is its attempt to combine theory and policy in such a way that the analytic foundation of policy evaluation work is improved (Bandara, 1991:9).
2.4. General application of AGE models
Although the foundations for AGE models were already laid in the 1960s, applied general models have been only been constructed and implemented from the early 1970s (Bandara, 1991:10). There are three main factors contributing to why AGE models only came to be implemented effectively
since the 1970s. First, the relative importance of the sectoral dimension of economic growth was insignificant in the phase of stable conditions that most countries experienced during the 1960s; second, the development of suitable solution algorithms and affordable software and hardware facilities for the solving of large sets of simultaneous equations were limited up to the early
1970s; and finally, more attention was paid to linear programming models with emphasis on resource allocation (Bandara, 1991:10).
All applied general equilibrium models share three common features, the first being that they focus on the real side of the economy. Second, their supply and demand functions explicitly reflect the behaviour of profit maximising producers and utility maximising consumers. Third, both quantities and relative prices endogenous to these models, as well as the resource allocation patterns determined by them, have their roots in the Walrasian general equilibrium. (Weintraub, 1977:8) Despite these common features, all these models differ from each other with respect to detailed specifications and solution algorithms (Shoven and Whalley, 1992:71).
AGE modelling has been widely used by policy analysts in addressing contemporary policy issues in mixed economies since the 1970s in the areas of trade policy, income distribution, issues related to external shocks and structural adjustments, government fiscal policy related issues and choice of development strategies such as long-term growth and structural changes (Bandara, 1991:13). However, there is still considerable debate in the economic profession regarding the value and appropriateness of using AGE
models for policy analysis. The following points of advantages as well as criticism of AGE models are brought for and against the use of AGE models based on Bandara (1991:29-31) and Borges (1986:15-22), grouped by the author broadly into areas related to methodological, model application and policy application aspects:
2.4.1. Methodological advantages
a) Solid microeconomic foundations: Probably the most important
strength of the general equilibrium methodology is its solid microeconomic foundation. A typical AGE model specifies the behaviour of all the economic agents using widely-accepted principles of optimisation and choice, which are operational and remain the most frequently used basis for empirical work.
b) Standard methods allows for no ad hoc specification resulting in
transparent model structures: AGE models use standard methods to
describe all the relationships among variables resulting in the fact that any ad hoc specification is precluded and making the structure of such models more transparent.
c) Embodied theory provides validity checks on results. The embodied theory also provides a precise check on the validity of the results, since it is impossible (unless the model contains errors in specification) that a model will lead to results which are contrary to what the underlying theory predicts.
d) Coherent theoretical foundation results in internal consistency: Related to the advantages of a coherent theoretical foundation is the
issue of internal consistency. AGE models by nature are internally consistent.
e) Makes possible the simulation of complex interrelationships. AGE models allow the analyst to integrate into a single structure a whole series of effects which could not possibly be accounted for coherently in an informal manner therefore making possible the simulation of complex interrelationships while clarifying the role and impact of different factors.
f) Can handle large policy shocks: The fact that AGE models are solved numerically and not analytically allows for AGE models to handle large policy shocks since it does not depend on the assumptions of small change.
2.4.2. Policy applications advantages
g) Provides framework for understanding and analyzing policy questions: AGE models are capable of providing useful insight into
important policy problems, by presenting an especially useful framework for understanding and managing structural change.
h) Provides quantitative significance of the mechanisms analyzed by
theory as well as indirect effects of policy: In policy debate and
formulation, more is needed than the qualitative insights that pure theory can yield. Knowledge of the quantitative significance of the various mechanisms analyzed by theory is also required, including the
indirect effects of policies which may escape intuition and thus the attention of theorists.
i) Allows for highly disaggregated models allowing structural impact analysis: The flexibility of the solution algorithms has made
possible the development of highly disaggregated models, which also contributes to their practical usefulness. It is well known that many policy actions or exogenous shocks will have an overall impact on the economy which is much smaller than their effect on the structure of the economy.
j) Provides the possibility of deriving better measures of the welfare
gain or loss: AGE models provide the possibility of deriving better
measures of the welfare gain or loss associated with a new policy. Very often the impact of a policy decision is discussed in terms of very imperfect measures of welfare- such as income or gross national product. It is necessary to go beyond those rough estimates of the gains and losses associated with the change.
k) Provides a bridge between the theorist, the planner and the
practical policy maker: AGE models provide a bridge between the
theorist, the planner and the practical policy maker. Policy makers will be able to recognize some of the real-world dilemmas they face in the questions addressed by the models, while theorists will recognize the fundamental theoretical structure and will be able to relate the functioning of AGE models to known theorems and analytical results.
2.4.3. Methodological disadvantages
a) Over reliance on unrealistic neo-classical assumptions: AGE
perfectly competitive markets and constant returns to scale; in practice the situation of persistent disequilibrium results in additional costs that may be incurred as the economy continues operating without all markets clearing.
b) Data and parameter requirements: The data and parameter values used in the models are often criticized as being inadequate.
2.4.4. Policy applications disadvantages
c) Assumption of general equilibrium not relevant for all policy
issues: The general equilibrium assumption rules out the use of these
models for many important policy decisions. In addition it is assumed that nothing happens until equilibrium is reached, precluding the analysis of the impact of certain policies on the path of change. The general equilibrium approach is therefore not very relevant to e.g. the discussion of macroeconomic issues related to stabilization policy. d) AGE results are difficult to convey to policymakers in practice: A
major criticism with regard to practical limitations is that the calculated results from AGE models are difficult to explain to policy makers.
2.4.5. Model application disadvantages
e) Lack of empirical validation of the models: One of the most frequently mentioned weakness is the lack of empirical validation of the models, in the sense that usually there is no measure of the degree to which the model fits the data or tracks the historical facts. The models cannot be used to replicate the evolution of the economy in the past as a means of checking their validity.
f) Does not deal with adjustment processes and associated costs:
The treatment of inter-temporal issues - saving and investment decisions - and of expectations. The general equilibrium approach is directed towards long-term questions. Its results should be interpreted in that context. Traditional general equilibrium models do not deal with adjustment processes and the costs associated with them, along the path between today's situation - or the "base case" equilibrium - and the new scenario - often called the "revise case".
g) The treatment of technological progress in the long term context: Given that these models are designed to look at long term issues it is somewhat contradictory that their structure does not include a more careful treatment of technological change, the implications of which can be far reaching in the long run.
h) Unrealistic assumption of expenditures equalling revenues: A standard assumption of the first generation of general equilibrium models that all economic agents are on their budget constraints, that is, expenditures equal revenues. Most important consequences of this approach are that the government runs a balanced budget and that there are no international current account imbalances. This however is not the case in more recent applications of AGE models.
i) Lack of explicit financial market treatment: Since all economic agents spend all their incomes in the purchase of goods and services, there is no scope for financial markets. Consequently, almost all of the existing models do not include money or monetary assets. Therefore, the models have been used to study resource allocation issues, but not
financial or monetary policy issues. In particular, it would be impossible to discuss problems related to inflation in the context of the existing models.
j) Simplistic treatment of foreign sector: Most existing general equilibrium models have a very inadequate treatment of the foreign sector, and in particular of net trade flows. Again this criticism is not necessarily valid for certain modelling initiatives such as that of the Global Trade Analysis Project (GTAP) coordinated by the Centre for Global Trade Analysis, at the Department of Agricultural Economics at Purdue University.
k) Complexities of relevant closure selection: When closing the model - that is, when specifying the last element of it - all degrees of freedom are lost and any behavioural relationships must be consistent with what has been specified for the rest of the model. Naturally, since the last element of the model has this "residual" nature, it must be a relatively unimportant component, given the application in question.
Despite the above mentioned criticisms AGE models are capable of providing useful insight into important policy problems, by presenting an especially useful framework for understanding and managing structural change. In policy debate and formulation, more is needed than the qualitative insights that pure theory can yield. Knowledge of the quantitative significance of the various mechanisms analyzed by theory is also required, including the indirect effects of policies which may escape intuition and thus the attention of theorists. AGE models provide a bridge between the theorist, the planner and the
practical policy maker. Policy makers will be able to recognize some of the real-world dilemmas they face in the questions addressed by the models, while theorists will recognize the fundamental theoretical structure and will be able to relate the functioning of AGE models to known theorems and analytical results.
Many quantitative methods are at the disposal of policy makers for assessing economic impacts of projects or events; however, the most commonly accepted is input-output analysis (Lundberg et al. 1995). The basic building blocks for AGE models from an information point of view are based on input-output tables, extended to social account matrices. These provide the basis for a AGE's information requirements. The relationship between these two elements and an AGE model is discussed in the next section.
2.5. Basic building blocks for AGE models
2.5.1. Input-output analysis
Input-output analysis is an analytical framework with the fundamental purpose to analyse the interdependence of industries in an economy (O'Connor and Henry, 1975:1). An input-output model in its most basic form consists of a system of linear equations, each one of which describes the distribution of an industry's product throughout the economy. While input-output analysis is a widely accepted and useful means of economic impact analysis, it is limited in that it does not reveal the personal income distribution effects across different household income segments (Holland and Wythe, 1993) and gives no consideration to industry occupation, skills and wages and the resulting
income effects. Therefore, using input-output models to assess economic impacts will not allow for a clear picture of which household income groups are benefiting / suffering and which are not, from the specified issue under review.
2.5.2. Social accounting matrices
A social accounting matrix (SAM) is a data system, including both social and economic data for an economy. The data sources for a SAM come from input-output tables, national income statistics, and household income and expenditure statistics (Cameron, 2003). Therefore, a SAM is broader than an input-output table and typical national accounts, showing more detail about all kinds of transactions within an economy. However, an input-output table records economic transactions irrespective of the social background of the transactors. A SAM, contrary to national accounts, "... attempts to classify various institutions to their socio-economic backgrounds instead of their economic or functional activities" (Chowdhury and Kirkpatrick, 1994:58).
A SAM is a way of logical arrangement of statistical information, concerning income flows in a country's economy within a particular time period (usually a year). It can provide a conceptual basis to analyse both distributional and growth issues within a single framework (Statistics South Africa, 1998a:7). For
instance, a SAM shows the distribution of factor incomes of both domestic and foreign origin, over institutional classes and re-distribution of income over these classes. In addition, it shows the expenditure of these classes on consumption, investment and savings made by them.
King (1988) points out that a SAM has two main objectives:
a) first, organising information about the economic and social structure of a country over a period of time; and
b) second, providing statistical basis for the creation of a plausible model capable of presenting a static image of the economy along with simulating the effects of policy interventions in the economy or other economic impacts.
A SAM, coupled with a conceptual framework that contains the behavioural and technical relationships among variables within and among sets of accounts, can be used for the evaluation of the economy-wide effects of policy changes or other economic impacts rather than only for purely diagnostic purposes (Pyatt, 1988:349). The aforementioned conceptual framework is supplied in the form of an AGE model.
2.6. The AGE framework
The aim of AGE modelling is to convert the abstract representation of an economy into realistic, solvable models of actual economies. In the AGE framework the main focus of analysis is quantitative and is based on the empirical data from a particular country being investigated. One of the major features of AGE modelling is its attempt to combine theory and policy in such a way that the analytic foundations of policy evaluation work are improved. An AGE model can accordingly be described as an economy-wide model that includes feedback between demand, income and production structure, and where all prices adjust until decisions made in production are consistent with decisions made in demand (Dervis etal. 1985:132).
A typical AGE model has a theoretical structure that consists of equations describing, for some time period (Wing, 2004:5-8):
• producers' demands for produced inputs and primary factors; • producers' supplies of commodities;
• demands for inputs to capital formation; • household demands;
• export demands; • government demands;
• the relationship of basic values to production costs and to purchasers' prices;
• market-clearing conditions for commodities and primary factors; and • numerous macroeconomic variables and price indices.
Demand and supply equations for private-sector agents are derived from the solutions to the optimisation problems (cost minimisation and utility maximisation) which are assumed to underlie the behaviour of the agents in conventional neoclassical microeconomics. The agents are assumed to be price takers, with producers operating in competitive markets, which prevent the earning of pure profits (Bandara, 1991:12). In addition to this static core, the AGE model can include several accumulation relationships that link the values of stocks (capital by industry and net foreign debt) over time, relating them to initial conditions and to the values of the relevant accumulating flows (investment and depreciation by industry, and foreign borrowing).
The main equations of these models are derived from the constrained optimisation of neo-classical production and utility functions. Producers select inputs in order to minimise costs of a given output subject to non-increasing returns to scale industry production functions. At the same time consumers are assumed to select purchases in order to maximise utility functions subject to their budget constraints. Production factors are paid according to their marginal productivity. The government sector is included and imperfect competition can be introduced via price fixing, rationing and quantitative restrictions. At the equilibrium level these models' solutions provide a set of prices that ensures that all commodity and factor markets are cleared as well as making all individual agents' optimisations feasible and mutually consistent (Bandara, 1991:12).
The diagram in Figure 1 illustrates the simplified relationship between an IO table, SAM and AGE model:
Figure 1: Simplified relationship between and IO table, SAM and AGE Model
^ ^ W e G e n e r a I l g T J | ^ ^
One can therefore summarise the relationship and context between 10 tables, SAMs and AGE models as follows (Cameron, 2003):
a) Input-output (10) analysis is an analytical framework with the fundamental purpose to analyse the interdependence of industries in an economy. An input-output table records economic transactions irrespective of the social background of the transactors.
b) A social accounting matrix (SAM) comes from input-output tables, national income statistics, and household income and expenditure statistics. Therefore, a SAM is broader than an input-output table and typical national accounts, showing more detail about all kinds of transactions within an economy.
c) An applied general equilibrium (AGE) model comes from a SAM, coupled with a conceptual framework that contains the behavioural and technical relationships among variables within and among sets of accounts. The aim of AGE modelling is to convert the abstract representation of an economy into realistic, solvable models of actual economies. In brief one can therefore state that an AGE model has the benefit that it can be used for a more detailed and realistic evaluation of the economy-wide effects of policy changes or other economic impacts than either an 10 analysis or SAM.
2.7. Closure rules for AGE models
The concept of "closure rules" refers to the way in which the number of endogenous variables in a model is set equal to the number of equations of the model in order to ensure that there will be a mathematical solution to the system of equations (or model) (Robinson, 1991; Whalley and Yeung, 1984).
characteristics, and it is therefore important to apply the correct closure depending on the specific question being analysed (Bandara, 1991:17). According to Bandara (1991:17) closure rules are generally classified into four basic types, namely:
a) The Keynesian closure - the assumption of full employment of production factors is relaxed with the implication that disequilibrium can occur in the labour market.
b) The Kaldorian closure - the assumption that wages of production factors must equal their marginal productivity is relaxed.
c) The Johansen closure - the assumption that household consumption is a function of only household income is relaxed and it is stated that full employment can be reached by shifting consumption.
d) The Classical closure - the assumption that investment is a constant is relaxed and investment is viewed as endogenous.
2.8. Interpretation of AGE results
AGE models are used mainly for comparative static analyses. This form of analysis involves comparing equilibrium positions as opposed to examining the path which the market follows when moving from the old to the new equilibrium. Therefore, the change in a system from one position of equilibrium to another is investigated, without regard to the transitional process involved in the adjustment (Zamagni, 1987:98). Results of comparative-static AGEs all refer implicitly to the economy at some future time period. To enable the reader to understand how to interpret these results a brief example is in order.
Figure 2: Stylised representation of change reported in AGE model results Employment C / Change B A 0 T
Source: Horridge etal. 1993 : 76
The example is illustrated by Figure 2, which graphs the values of for instance, employment, against time. The level of employment in the base period (period 0) is represented by A and C is the level which it would attain in T years time if some policy, for instance a tariff reform scheme, is implemented, all other things being equal. B represents the level which would be attained if the policy was not implemented. A comparative static AGE model would generate the percentage change in employment ((C-B)/B) x 100, illustrating how employment in period T would be influenced by the tariff policy in isolation. For these kind of simulations capital stocks are usually held at their pre-shock levels and econometric evidence suggests that a short-run equilibrium will be reached in about two years (T=2) (Horridge, et .at., 1993:73). Note that nothing is said about the path of change (the curved lines connecting A with levels B and C), only the change from B to C.
Further developments in the AGE field make the inclusion of dynamic elements, such as stock/flow accumulation relations between capital stocks and investment as well as between foreign debt and trade deficits, possible. The ORANI-F model (See Horridge, et al., 1993 for a detailed description of the ORANI-F model) is an example of an AGE model with dynamic capability. ORANI-F is used for making medium-term forecasts for the Australian economy.
The main source of data for any AGE model is normally supplied in the form of a social accounting matrix (SAM) as explained in the previous section. Since a SAM represents the circular flow of income in a country, it is considered to be a natural candidate for a AGE to be based upon (Pyatt, 1988). According to Pyatt (1988) any model based on a SAM will also satisfy Walras' law. The relationship between SAMs and models can be described as asymmetric in the sense that for each model there exists a corresponding SAM, but for any given SAM a variety of possible models exist.
2.9. Summary
This chapter illustrated the need for AGE models as tools for policy makers in South Africa through the fact that economic models (and more specifically AGE models) can provide key insights for analysis of comprehensive packages of economic and non-economic policy instruments within a consistent framework and insights generated by this modelling can help in evaluating various policy options.
The concept of general equilibrium in the context of AGE models as well as Walras' law, a central building block of general equilibrium, was explained, as well as developments of the application of AGE models from the 1960's onwards. Common features as well as constraints of AGE models were discussed, after which the basic building blocks of information for AGE models, namely input-output tables and social accounting matrices, were discussed in brief.
The aim of AGE modelling, to convert the abstract representation of an economy into realistic, solvable models of actual economies with the main focus of analysis being quantitative based on empirical data was highlighted. The concept of closure rules was explained and the fact that it is important to apply the correct closure depending on the specific question being analysed was highlighted. Lastly a short explanation on the interpretation of percentage change results from a typical linearised solution approach was explained. In South Africa, the potential benefits of using AGE modelling was realised in the early 1990s, following the availability of SAMs for the economy. The development and use of AGEs in South Africa is discussed in the next chapter.
CHAPTER 3: APPLIED GENERAL EQUILIBRIUM MODELLING IN SOUTH AFRICA
3.1. Introduction
In the previous section, an overview was given of the theoretical foundations of AGE modelling. Although the first SAM for South Africa was published in 1986 by the then Central Economic Advisory Services, the SAM was for 1978 and excluded useful household income and race dimensions (SEAD, 1986). In 1993 a provisional SAM was published by the Central Statistical Services (with base year 1988 for economic variables and 1991 Census for household variables) for the South African economy, and was used in descriptive manner to analyse questions relating to the interrelationships between growth, distribution of income and production in the economy (SSD, 1993). The first AGE models to make use of this SAM framework, however only dates from the early 1990s.
This chapter will discuss the development of AGE modelling in South Africa, and will show that despite a significant growth in the number of models and applications (both for policy as well as for private firm use) after 15 years of AGE modelling there is still a need to go beyond the country level to sub-national AGE models.
The chapter is structured as follows. Section 3.2 provides a chronology of applications of AGE models in South Africa since the early 1990's. Section 3.3 provides an overview of the focus of major applications of AGE models in South Africa over this period, illustrating the lack of sub-national detail and
applications for South Africa over this period. Section 3.4 summarises the chapter.
3.2. Major applications of AGE models in South Africa
The chronology presented in Table 1 is based on Rossouw (2004)3 in which
they provide and overview of the developments in AGE modelling in South Africa since the 1990s, as well as the author's own experience4 and
involvement in these events.
Rather than summarising the analysis of Rossouw (2004) here, this section will present a chronological timeline to represent the developments highlighted by Rossouw (2004), as well as to expand the list with some additional entries based on the authors' research as well as personal involvement in some of these efforts.
The chronology is non-exhaustive since the specific developments listed excluded works non-academic in nature. Over the period discussed in this chronology AGE models were applied in various impact studies of non-academic and non-public nature, some of which the author was part of personally. In addition, applications of multi-country model developments in which South African issues were analysed are also excluded. An example is
3 The author added some information as well as corrected some date references.
4 The author was part of the initial AGE modelling team at the Industrial Development
Corporation of South Africa during the period 1994 - 1996. During this period the IDCGEM
was applied to study a variety of topics, but the focus was mostly on international trade
the Global Trade Analysis Project (GTAP) coordinated by the Centre for Global Trade Analysis, at the Department of Agricultural Economics at Purdue University.
As mentioned by Rossouw (2004), the Industrial Development Corporation Computable General Equilibrium Model (IDCGEM), based on the Monash ORANI approach, dominated the AGE modelling scene during the 1990s. However, due to difficulties experienced by the IDC in retaining the capacity to utilise the model, its dominance of AGE modelling in South Africa started to diminish and the new millennium saw the emergence of new players in the field, both local and international, along with the appearance of a number of alternative modelling strategies.
One reason for a fresh wave of interest in AGE modelling in South Africa since 2000 was due to the World Bank's interest in the implications of HIV/Aids on the South African economy (Arndt and Lewis, 2000). The momentum was picked-up and sustained by researchers under the auspices of the International Food Policy Research Institute (IFPRI) (Thurlow et a/. 2002 & 2003), but academic institution modellers have also been playing an active role in building new models to answer to new policy challenges (McDonald & Punt, 2005:88).
Table 1: Chronological developments of AGE modelling in SA 1993-2005
taaa 1984 1995 me 1497
(1)"*Potchefstroom University and the University of Copenhagen Naude & Bhxen | [ I Model = Provisional Multi-sector CGE | Major focus = Government sector dominance in SA economy R - no regional aspects I
I I I I
( 2 ) * Industrial Development Corporation (IOC) of South Africa Joubert et. al supported by Horridge et. al Model = 96-sector CGE based on ORANI-F Major focus = Free Trade - EUFTA and GATT R- no regional aspects I
(3)*National Economic Policy unit (NIEP) Gelb et.al supported by Brain et. al
" Model = Multi-sector macroeconometric model / eclectic CGE Major focus = Reconstruction & Development Programme (RDP) R - no regional aspects I I I (4)■♦Industrial Development Corporation (IDC) of South Africa
De Jongh, Cameron et.al supported by Horridge ef. al Model = 103-sector CGE based on ORANI-F | Major focus = Government Expenditure & Provincial Implications R - Top-down Provincial (9) share) j
Major focus = Government Expenditure - Provincial GOP implications
1 ! I
(5)*Development Bank of Southern Africa | van Seventer et.al supported by Gibson ef. a/ Model = Structuralist CGE | Major focus = Impact of new ANC government policies R - no regional aspects I I (6)^Industrial Development Corporation (IDC) of South Africa
Cameron et.al supported by Horridge ef. al I Model = 103-sector CGE based on ORANI-F | Major focus = Government Expenditure & RDP implications R- Top-down Provincial (9) share I I
2 0 M _ _ ^ 0 1 . _ 2003 (9)-»Centre on Globalisation & Sustainable Development
Humphreys | | Model = Small Open Economy 3-sector CGE Major focus = Trade Liberalisation & Poverty R - no regional aspects
(10)-»WorldBank Arndt S. Lewis
Model = Solow-type standard Neo-classical CGE Major focus = Impact of HIV/Aids
R -no regional aspects I
I
*(11)-*-WorldBank | Devarajan &van der Mensbrugghe Model = Standard Neo-classical CGE
Major focus = Removal of Trade Barriers impact on housed R - no regional aspects
(12)*University of Pretoria DeWetef.a/. Model = ORANI-G for SA
Major focus = Fiscal reform with focus on Coal mining sector R - no regional aspects
2004 2006
O j ^ Jmvcio.ty of Pretoria Van Heerden etal.
Model =UPGEM-based on ORANI-G Major focus = Energy & Greenhouse Gas emissions R - no regional aspects I (21)->lndustrial Development Corporation (IDC) of South Africa
Caefzee et.al supported by Naude et. al I Model = 103-sector CGE based on ORANI-F [ Major focus = Causes and extent of inequality in South Africa R - Top-down Provincial (9) share
*(22)*University of Pretoria
Mabugu et.al. \ Model = fntertemporal dynamic CGE
Major focus = Draft-transitional changes of trade policy reform R - no regional aspects
(13)*Unh/ersity of Pretoria Kearny et.al.
Model = Stylized Walrasian Closed-economy model Major focus = Fiscal policy with focus on Value Added Tax R- no regional aspects I I (14)-*Dept of Agriculture - Elsenburg Western Cape
Punt & McDonald | I Model = Neo-classical CGE | Major focus = Exports & provincial Agriculture sectors R - Provincial (Western Cape) SAM & CGE (7)^Industrial Development Corporation (IDC) of South Africa
Coetzee et.al supported by Naude ef al I Model = 103-sector CGE based on ORANI-F | Major focus = Accelerated Tariff Reduction Programme, Currency Depreciation R-Top-down Provincial (9) share I I I
l
*(8)-»ndustrial Development Corporation (IDC) of South Africa Coetzee et.al supported by Naude et. al I Model = Dynamic 103-sector CGE based on ORANI-F | Major focus = Sectoral Prospects - 1997 to 2001 projections R - Top-down Provincial (9) share
(17)*lnternatianal Food Policy Research Institute (IFPRI) Thurlow et.al. \ Model = Standard Neo-classical CGE Major focus = HIV/Aids S. Trade Policy R - no regional aspects I
'(23)■♦University of North West Naude et.al. \ Model =DTIGEM- based on ORANI-G
Major focus = Transport sector productivity improvements R - no regional aspects
(18)^Deptof Agriculture, Western Cape Pauw and Edwards ( Model = Standard Neo-classical CGE
Major focus = Adapting IFPRI - focus on semi & law skilled wages R - no regional aspects
** (15)*lnternational Food Policy Research Institute (IFPRI) Thurlow et.al. | l Major focus = Standardising IFPRI for SA
R - no regional aspects ( [
l I I
(16)*lnternational Food Policy Research Institute (IFPRI) Thurlow et.al. | I Model = Dynamic Neo-classical CGE | Major focus = Various Poverty & Trade Policy related R - no regional aspects I I
(19) ■♦University of Stelenbosch Van Schoor and Burrows )
Model = Standard Neo-classical CGE based on IFPRI model
Major focus = Adapting IFPRI - focus on imperfect competition & returns to scale R - no regional aspects I I I
